Many older adults respond to high out-of-pocket drug costs by skipping, reducing dosages of or not filling prescriptions for their medications. A primary goal of the new Medicare drug benefit (Part D) was to increase the availability of drug coverage and thereby improve medication use among older adults. Medicare Part D may have a disproportionate impact on medication use among beneficiaries in minority groups. For instance, African American beneficiaries have enrolled in Part D in greater numbers (62% vs. 48%) than Caucasian beneficiaries. Moreover, African American beneficiaries are more likely than Caucasian beneficiaries to qualify for generous Part D low-income subsidies. As a result, Part D may help to alleviate racial/ethnic disparities in clinical processes of care and health outcomes that have been documented for a range of chronic conditions among older adults. We will assess the impact of Part D on racial disparities in medication use and health outcomes for older adults with diabetes, a chronic condition with increasing prevalence. Given the worse glycemic control and higher diabetes-related mortality among African Americans compared to Caucasians, our proposed investigations have critical implications for health disparities. The long-term objective of our proposed study is to reduce racial/ethnic disparities in health by examining the impact of changes in insurance coverage among the elderly. This proposal has two primary objectives. First, we will assess racial differences in antidiabetic medication use, glycemic control, and diabetes-related hospitalizations and mortality for an elderly cohort with diabetes before the implementation of Part D. We hypothesize that older African American beneficiaries with diabetes will have lower levels of antidiabetic medication use, worse glycemic control, and higher diabetes-related hospitalization and mortality rates compared to Caucasian beneficiaries. Second, we will examine whether Part D mitigates racial disparities in medication use, glycemic control, and diabetes-related hospitalizations and mortality in the elderly. We hypothesize that older African Americans compared to Caucasians will experience a larger improvement in medication use, glycemic control, and diabetes-related hospitalizations and mortality. Our study will use longitudinal data from an 11-year, Medicare population-based prospective cohort study. We will adjust our estimates of the effect of race and Part D for socioeconomic status, comorbidities, health, functional and cognitive status, access to care, and other factors. Our research has important policy implications. Medicare Part D -- the most substantial expansion of insurance benefits to older adults since the program's inception over 40 years ago -- is projected to cost just under $1 trillion between 2007 and 2017. If Medicare Part D reduces racial disparities in medication access, our study will inform other efforts to expand insurance coverage to vulnerable populations. Our multidisciplinary research team has extensive experience conducting health policy analyses, pharmacoepidemiologic studies with geriatric populations, health disparities and diabetes research, and is uniquely positioned to shed light on this critical policy issue.